Adaptable incident surveillance system

Abstract

A camera orientation control system for changing one or more camera orientations of one or more cameras towards a potential incident that is not currently in a field of view of the one or more cameras. The camera control system processes image frames for head orientation extraction, determines an amount of synchronous alignment in a proportion of head objects of the head objects to computationally assess that they are towards a new direction, and controls at least one camera to move the field of view of at least one camera of the one or more cameras towards the new direction. The system is useful for incident detection and extending or augmenting the capabilities of CCTV infrastructure.

The invention claimed is: 
     
       1. A camera orientation monitoring system for changing one or more camera orientations of one or more cameras towards a potential incident that is not currently in a field of view of the one or more cameras, the camera orientation monitoring system comprising:
 a processor coupled to a computer memory and non-transitory computer readable media, the processor configured to:
 process a plurality of image frames recorded by at least one camera of the one or more cameras to extract one or more head orientations corresponding to head objects identified for a plurality of individuals in the plurality of image frames over a first duration of time represented in the plurality of image frames; 
 determine, from the one or more extracted head orientations, an amount of synchronous alignment in a proportion of head objects of the head objects in a new direction within a threshold range of angles; 
 based at least on the amount of synchronous alignment in the head objects in the new direction, control at least one camera of the one or more cameras to move the field of view of at least one camera of the one or more cameras towards the new direction; 
 
 wherein the amount of synchronous alignment is based on a balanced weighting of individual contributions corresponding to each head object to a synchronous alignment score; and wherein the control of the at least one camera occurs when the synchronous alignment score is greater than a threshold synchronous alignment score. 
 
     
     
       2. The camera orientation monitoring system of  claim 1 , wherein the balanced weighting includes applying larger weightings for heads that are closer to an inferred incident. 
     
     
       3. The camera orientation monitoring system of  claim 1 , wherein the processing of the plurality of image frames includes extracting one or more head positions each corresponding to a corresponding head object of the head objects, and the head positions are utilized to modify the balanced weighting based on their positioning relative to an inferred accident. 
     
     
       4. The camera orientation monitoring system of  claim 3 , wherein the new direction is used to establish an axis, and the balanced weighting for each head object is weighted based on positioning relative to the axis of the new direction. 
     
     
       5. The camera orientation monitoring system of  claim 3 , wherein the one or more head positions and the one or more head orientations of the proportion of heads that have oriented towards the new direction are used together to establish an intersection point, and balanced weighting for each head object is weighted based on positioning relative to the intersection point. 
     
     
       6. The camera orientation monitoring system of  claim 5 , wherein upon a determination that the intersection point is outside of an available field of view capable of being covered by the one or more cameras, an alert notification is generated. 
     
     
       7. The camera orientation monitoring system of  claim 1 , wherein the proportion of the head objects required for the synchronous alignment is selected from the group of percentages of at least one of 30%, 40%, 50%, 60%, 70%, 80%, 90%, or 100% of the head objects. 
     
     
       8. The camera orientation monitoring system of  claim 1 , wherein the one or more cameras include a rotating camera that operates on a pre-programmed routine that is overridden by the control to move the field of view of the at least one camera of the one or more cameras towards the new direction. 
     
     
       9. The camera orientation monitoring system of  claim 8 , wherein the one or more cameras includes two cameras, the rotating camera and a fixed camera having a fixed field of view. 
     
     
       10. The camera orientation monitoring system of  claim 8 , wherein the control to move the field of view is provided as signal inputs to actuators of the rotating camera. 
     
     
       11. A method for changing one or more camera orientations of one or more cameras towards a potential incident that is not currently in a field of view of the one or more cameras, the method comprising:
 processing a plurality of image frames recorded by at least one camera of the one or more cameras to extract one or more head orientations corresponding to head objects identified for a plurality of individuals in the plurality of image frames over a first duration of time represented in the plurality of image frames; 
 determining, from the one or more extracted head orientations, an amount of synchronous alignment in a proportion of head objects of the head objects in a new direction within a threshold range of angles; 
 based at least on the amount of synchronous alignment in the head objects in the new direction, controlling at least one camera of the one or more cameras to move the field of view of at least one camera of the one or more cameras towards the new direction; 
 wherein the amount of synchronous alignment is based on a balanced weighting of individual contributions corresponding to each head object to a synchronous alignment score; and wherein the control of the at least one camera occurs when the synchronous alignment score is greater than a threshold synchronous alignment score. 
 
     
     
       12. The method of  claim 11 , wherein the balanced weighting includes applying larger weightings for heads that are closer to an inferred incident. 
     
     
       13. The method of  claim 11 , wherein the processing of the plurality of image frames includes extracting one or more head positions each corresponding to a corresponding head object of the head objects, and the head positions are utilized to modify the balanced weighting based on their positioning relative to an inferred accident. 
     
     
       14. The method of  claim 13 , wherein the new direction is used to establish an axis, and the balanced weighting for each head object is weighted based on positioning relative to the axis of the new direction. 
     
     
       15. The method of  claim 13 , wherein the one or more head positions and the one or more head orientations of the proportion of heads that have oriented towards the new direction are used together to establish an intersection point, and balanced weighting for each head object is weighted based on positioning relative to the intersection point. 
     
     
       16. The method of  claim 15 , wherein upon a determination that the intersection point is outside of an available field of view capable of being covered by the one or more cameras, an alert notification is generated. 
     
     
       17. The method of  claim 11 , wherein the proportion of the head objects required for the synchronous alignment is selected from the group of percentages of at least one of 30%, 40%, 50%, 60%, 70%, 80%, 90%, or 100% of the head objects. 
     
     
       18. The method of  claim 11 , wherein the one or more cameras include a rotating camera that operates on a pre-programmed routine that is overridden by the control to move the field of view of the at least one camera of the one or more cameras towards the new direction. 
     
     
       19. The method of  claim 18 , wherein the one or more cameras includes two cameras, the rotating camera and a fixed camera having a fixed field of view. 
     
     
       20. A non-transitory computer readable medium, storing machine interpretable instruction sets, which when executed by a processor, cause the processor to perform a method for changing one or more camera orientations of one or more cameras towards a potential incident that is not currently in a field of view of the one or more cameras, the method comprising:
 processing a plurality of image frames recorded by at least one camera of the one or more cameras to extract one or more head orientations corresponding to head objects identified for a plurality of individuals in the plurality of image frames over a first duration of time represented in the plurality of image frames; 
 determining, from the one or more extracted head orientations, an amount of synchronous alignment in a proportion of head objects of the head objects in a new direction within a threshold range of angles; 
 based at least on the amount of synchronous alignment in the head objects in the new direction, controlling at least one camera of the one or more cameras to move the field of view of at least one camera of the one or more cameras towards the new direction; 
 wherein the amount of synchronous alignment is based on a balanced weighting of individual contributions corresponding to each head object to a synchronous alignment score; and wherein the control of the at least one camera occurs when the synchronous alignment score is greater than a threshold synchronous alignment score.